2 Releases published by 1 person

• v2.4.11 DSPy v2.4.11

  published Jul 6, 2024

• v2.4.12 DSPy v2.4.12

  published Jul 8, 2024

22 Pull requests merged by 18 people

• docs(readme): fix minor typo

  #1320 merged Jul 26, 2024

• databricks_rm: returning extra columns

  #1313 merged Jul 24, 2024

• add base_url alias for OpenAI LM

  #1281 merged Jul 23, 2024

• fix(dependency/semver): updated semver to allow tests to run

  #1299 merged Jul 21, 2024

• Fix conditional in Module.named_parameters() to prevent infinite recursion

  #1294 merged Jul 20, 2024

• Add metric_threshold to MIPROv2

  #1296 merged Jul 20, 2024

• miprov2 patch for proper parsing

  #1280 merged Jul 16, 2024

• Update RunLLM's chat widget script path

  #1282 merged Jul 16, 2024

• Revert "cache:use temp dir by default"

  #1284 merged Jul 16, 2024

• cache:use temp dir by default

  #1278 merged Jul 16, 2024

• Add DSPy visualizer to the README

  #1277 merged Jul 13, 2024

• loosen joblib, backoff dependency restrictions

  #1267 merged Jul 12, 2024

• added configurable backoff max time

  #1268 merged Jul 10, 2024

• Enable Langchain Tool Execution via Avatar

  #1259 merged Jul 8, 2024

• 🐛 fix: http_request_kwargs in HFClientVLLM

  #1250 merged Jul 8, 2024

• fix(MiproV2): custom view_data_batch_size for dataset summary creation

  #1217 merged Jul 8, 2024

• Add filter for QdrantRM

  #1239 merged Jul 8, 2024

• fix: Support for Meta models provided by Amazon SageMaker

  #1241 merged Jul 8, 2024

• hotfix(dspy): update Together client self.kwargs to include "model"

  #1257 merged Jul 8, 2024

• Chat templates v2

  #1256 merged Jul 6, 2024

• Relax ruff requirements and patch import positions

  #1218 merged Jul 6, 2024

• Loosen Anthropic requirement

  #1226 merged Jul 4, 2024

9 Pull requests opened by 9 people

• feature(dspy): added compatibility with ANY llamaindex LLM

  #1225 opened Jul 1, 2024

• add LiteLLMVectorizer to allow more embeddings to be used

  #1240 opened Jul 3, 2024

• fix(mlc-llm): fix mlc-llm ChatModuleClient with MLCEngine

  #1285 opened Jul 16, 2024

• add OpenAICompatible

  #1287 opened Jul 17, 2024

• Add support for Langchain Tool use Optimization via Avatar

  #1289 opened Jul 17, 2024

• Support nested objects in set_attribute_by_name() via magicattr

  #1301 opened Jul 21, 2024

• fix (mipro): enable running mipro using only labeled demos

  #1305 opened Jul 22, 2024

• fix(dspy): added type check on evaluator result

  #1321 opened Jul 26, 2024

• feat(dspy): add system and format options to ollama

  #1323 opened Jul 26, 2024

39 Issues closed by 16 people

• Bootstrap temperature (0.7) overrides user temperature and LM temperature

  #1315 closed Jul 24, 2024

• MIPROv2 - Error generating fewshot examples: name 'metric_threshold' is not defined

  #1303 closed Jul 23, 2024

• Debug in DSPy

  #1304 closed Jul 23, 2024

• How to view the optimized prompt ？

  #1308 closed Jul 23, 2024

• Signatures rarely work out of the box

  #1275 closed Jul 21, 2024

• BuilderConfig doesn't have "trust_remote_code" key when loading HotPotQA

  #1288 closed Jul 21, 2024

• dspy.Retrieve cannot read csv file

  #1291 closed Jul 21, 2024

• semver dependency error when setting up dev container

  #1298 closed Jul 21, 2024

• import error: AttributeError: module 'threading' has no attribute '_Condition'. Did you mean: 'Condition'?

  #1297 closed Jul 20, 2024

• How to get multiple output from the dspy program? (e.g., two answers for a question)

  #1283 closed Jul 17, 2024

• Failed to parse JSON response from LLM served using vLLM

  #1242 closed Jul 15, 2024

• OpenAI + ReAct : TypeError: unhashable type: 'list'

  #1276 closed Jul 15, 2024

• question about the documentation of KNNFewShot

  #1262 closed Jul 12, 2024

• ModuleNotFoundError: No module named 'dspy.datasets'; 'dspy' is not a package

  #1273 closed Jul 12, 2024

• Tag releases

  #598 closed Jul 11, 2024

• Missing super class in gsm8k dataset class

  #1229 closed Jul 11, 2024

• Chat support in Ollama impacting normal generation, something with `"content": "".join(text)`

  #1270 closed Jul 11, 2024

• TypedChainOfThought and ChainOfThought have divergent reasoning output fields.

  #1235 closed Jul 9, 2024

• How to set the concurrency when using a custom api to access the model?

  #1236 closed Jul 9, 2024

• Does DSPY require cohere? Never saw this in the main documentation.....

  #1251 closed Jul 9, 2024

• Can we add query_filter filed for QdrantRM

  #1238 closed Jul 8, 2024

• lm.copy() not working for Together API

  #1254 closed Jul 8, 2024

• _call_model in aws_models.py raises error when using Sagemaker as aws_provider

  #1009 closed Jul 8, 2024

• HFClientVLLM is missing http_request_kwargs

  #1249 closed Jul 8, 2024

• GPT4T is hard coded

  #1258 closed Jul 8, 2024

• CoT and Tutorial not working anymore

  #1253 closed Jul 6, 2024

• How do I specify metadata query parameters in Retrievers?

  #1212 closed Jul 3, 2024

• QdrantRM hard coded payload name as 'document'

  #1237 closed Jul 3, 2024

• Missing import GSM8K dataset in __init__.py

  #1227 closed Jul 2, 2024

• How to use the dspy to get embeddings for MTEB tasks?

  #1220 closed Jun 30, 2024

• Refactor the function `set_attribute_by_name`

  #405 closed Jun 28, 2024

• How to use endpoint for model published on EC-2 server

  #416 closed Jun 28, 2024

• BayesianSignatureOptimizer - missing 1 required positional argument: 'example'

  #427 closed Jun 28, 2024

• how to connect to llama-2-7b.ggmlv3.q4_K_S.bin

  #438 closed Jun 28, 2024

• function calling

  #192 closed Jun 28, 2024

• Docstrings

  #328 closed Jun 28, 2024

• Support OpenAI Batch requests.

  #886 closed Jun 28, 2024

• Mipro crash in mipro_optimizer.py

  #865 closed Jun 28, 2024

• What castle did Thomas Gregory inherit?

  #997 closed Jun 28, 2024

38 Issues opened by 30 people

• How to handle Exceptions when calling Azure OpenAI model API

  #1326 opened Jul 26, 2024

• Integrating XML Validation Feedback into Optimization Processes

  #1325 opened Jul 26, 2024

• KeyError: 'max_tokens'.

  #1324 opened Jul 26, 2024

• Integrate dspy with qwen 72B model

  #1322 opened Jul 26, 2024

• MIPRO compile ERROR

  #1319 opened Jul 25, 2024

• LlamaIndexRM returning Example with attribute `text` instead of `long_text`

  #1318 opened Jul 25, 2024

• Default top_p=1 everywhere. Why?

  #1317 opened Jul 24, 2024

• COPRO: doc fix

  #1316 opened Jul 24, 2024

• Field descriptions for individual items in a Pydantic model

  #1314 opened Jul 24, 2024

• Improper response with multiple output fields

  #1312 opened Jul 24, 2024

• max_tokens error for google's models

  #1311 opened Jul 24, 2024

• ReAct Fails on minimal Examples

  #1310 opened Jul 24, 2024

• Using TypedChainOfThough caused reasoning to be replaced by output schema

  #1307 opened Jul 22, 2024

• Predictors with Ollama return input prompt in answer

  #1306 opened Jul 22, 2024

• `named_parameters()` and `named_submodules()` for dictionary attributes with non-string keys

  #1302 opened Jul 21, 2024

• dspyComponent module only allows for a dspy.Predict module as input

  #1300 opened Jul 21, 2024

• Miprov2 doesn't work with TypedPredictors

  #1295 opened Jul 19, 2024

• List index error using MIPROv2

  #1293 opened Jul 19, 2024

• dsp.settings to modify templates. Where can I find them?

  #1292 opened Jul 19, 2024

• 'ReAct' object has no attribute 'best_programs'

  #1290 opened Jul 18, 2024

• Deprecation warnings point to non-existing `dspy.set_log_level`

  #1286 opened Jul 16, 2024

• Question: BootstrapFewShot - Intermediate Label Bootstrapping

  #1274 opened Jul 12, 2024

• Claude does not work with the new experimental setting.

  #1271 opened Jul 11, 2024

• AWSMeta incorrectly raises max_tokens error

  #1269 opened Jul 11, 2024

• DSPy prgram timeout issue and streaming support

  #1266 opened Jul 10, 2024

• TypeError: metaclass conflict when importing dspy LangChainPredict

  #1265 opened Jul 10, 2024

• Feature: Use "Retry-After" Header when retrying requests instead of exponential backoff

  #1263 opened Jul 9, 2024

• AssertionError in Trial, failed with value None: `assert max_depth > 0`

  #1252 opened Jul 5, 2024

• Question: Assertions vs typed predictor + enum for multiple choice question

  #1247 opened Jul 5, 2024

• TypedPredictor：Too many retries trying to get the correct output format.

  #1246 opened Jul 5, 2024

• Question: How to get the token count of the prompt for maximum utilization of the context window during prompt compression of RAG pipeline?

  #1245 opened Jul 4, 2024

• Support custom signature format

  #1244 opened Jul 4, 2024

• [error ] Error for example in dev set: [dspy.evaluate.evaluate] filename=evaluate.py lineno=180

  #1243 opened Jul 4, 2024

• pyproject.toml / setup.py / requirements.txt disagreement

  #1234 opened Jul 2, 2024

• Extend generation and template.extract producing empty fields and malformed outputs.

  #1232 opened Jul 2, 2024

• Disable caching for individual calls

  #1231 opened Jul 1, 2024

• TypeError: '<' not supported between instances of 'int' and 'SignatureMeta'

  #1230 opened Jul 1, 2024

• I cannot save the model , pickle error

  #1221 opened Jun 29, 2024

32 Unresolved conversations

Sometimes conversations happen on old items that aren’t yet closed. Here is a list of all the Issues and Pull Requests with unresolved conversations.

• feat(dspy): integration of langfuse for full observability of LLMs apps

  #200 commented on Jul 1, 2024 • 1 new comment

• Fix completions endpoint crash

  #1196 commented on Jul 23, 2024 • 0 new comments

• feature(dspy): dspy-integration-with-langfuse

  #1186 commented on Jul 11, 2024 • 0 new comments

• Add Tests for extend generation logic

  #1184 commented on Jul 19, 2024 • 0 new comments

• Add support for Bing Seach

  #1107 commented on Jul 17, 2024 • 0 new comments

• feature(dsp) Added support for multimodality in googlevertexai client

  #855 commented on Jul 19, 2024 • 0 new comments

• Fix multiple openai clients

  #744 commented on Jul 4, 2024 • 0 new comments

• Add GPT-4 Vision API wrapper

  #682 commented on Jul 24, 2024 • 0 new comments

• misc(dspy): Move skycamp into example directiory

  #670 commented on Jul 12, 2024 • 0 new comments

• feat: Add support for async modules and async bootstrap

  #161 commented on Jul 27, 2024 • 0 new comments

• DatabricksRM retrieval using dspy.Retrieve() throws TypeError

  #1191 commented on Jul 26, 2024 • 0 new comments

• RAG.forward() missing 1 required positional argument: 'question'

  #1175 commented on Jul 24, 2024 • 0 new comments

• [Feature-Request] Support for GPT Vision

  #624 commented on Jul 20, 2024 • 0 new comments

• Default prompting strategy suboptimal?

  #409 commented on Jul 18, 2024 • 0 new comments

• TypeError: cannot pickle 'module' object - teleprompter.compile

  #1038 commented on Jul 18, 2024 • 0 new comments

• List index out of range error after passing long text in input fields

  #531 commented on Jul 18, 2024 • 0 new comments

• [Feature Request] Better Typed Predictors for Reliable JSON Generation

  #1001 commented on Jul 17, 2024 • 0 new comments

• No module named 'mlc_chat'

  #1206 commented on Jul 16, 2024 • 0 new comments

• DSPY Copro Tutorial for Hotpot QA doesn't work with Bedrock Claude Sonnet Model

  #1173 commented on Jul 16, 2024 • 0 new comments

• dspy and langfuse integration

  #1179 commented on Jul 14, 2024 • 0 new comments

• how to set dspy to make the result to be chinese or english?

  #1210 commented on Jul 13, 2024 • 0 new comments

• Triggering the Azure OpenAi (GPT 4) Triggers content warning and DSPy exception

  #1183 commented on Jul 12, 2024 • 0 new comments

• Dspy Assertions make execution sequential

  #1215 commented on Jul 12, 2024 • 0 new comments

• Cannot disable Caching

  #1203 commented on Jul 10, 2024 • 0 new comments

• OpenAI timeout

  #207 commented on Jul 9, 2024 • 0 new comments

• long_text attribute for retriever passages

  #166 commented on Jul 8, 2024 • 0 new comments

• [Request] - Adding functionality of other commercial LLMs ?

  #273 commented on Jul 8, 2024 • 0 new comments

• 3 bugs in one testcase: Hang, broken Retry on v2.4.9 and some other bonus bug on main

  #1080 commented on Jul 6, 2024 • 0 new comments

• How to use Function calling in DSPy?

  #1213 commented on Jul 5, 2024 • 0 new comments

• TypedChainOfThought throw ValueError: 'Too many retries trying to get the correct output format. Try simplifying the requirements.'

  #957 commented on Jul 2, 2024 • 0 new comments

• How to use dspy.OpenAI() like langchain.llms.OpenAI()

  #479 commented on Jul 1, 2024 • 0 new comments

• Wrong parsing of output fields

  #847 commented on Jun 28, 2024 • 0 new comments